From Baidu Encyclopedia, etc.
(A) normalized difference vegetation index (NDVI)
Calculating NDVI = (NIR-R) / (NIR + R), or the two reflectance bands
Detecting the state of growth of vegetation, the vegetation coverage and eliminate the portion of the radiation error and the like;
NDVI can reflect the background affect plant canopy, such as soil, wet ground, snow, leaves, roughness, etc., and is covered with vegetation related
A negative value indicates a cloud covered ground, water, snow, high reflectivity to visible light;
0 expressed bare soil rock or the like, and R is approximately equal to the NIR;
positive, indicating vegetation cover, and coverage increases with
Nonlinear way stretch contrast enhancement and NIR reflectance R is.
With respect to the one image and RVI are seeking you will find NDVI, RVI NDVI value increases faster than the rate of increase, i.e. having a lower sensitivity NDVI high vegetation area;
EVI commonly used in high LAI value, i.e., dense vegetation region;
values range from -1 to 1, the range is generally from 0.2 to 0.8 green vegetation region
Enhanced Vegetation Index (EVI) algorithm is one of the main topics of remote sensing data products algorithm biophysical parameters in the product, can reduce the impact of noise from the atmosphere and soil at the same time, reflect the situation stable vegetation measured. The vegetation index based on MODIS EVI having high spatial resolution, can reflect the characteristics of the vegetation in detail.
Red and near-infrared detection band set narrower range, not only improves the ability to detect the sparse vegetation, but also reduces the effect of moisture, while the introduction of the blue band and soil background scattering Aerosol were corrected.
① using image data analyzed by extracting vegetation change Vegetation Index;
according to the algorithm enhanced vegetation index by processing the noise from the atmosphere and soil, to produce EVI.tif
②EVI can describe differences within certain climatic zones of vegetation in different seasons. EVI uses to analyze vegetation change and climate change, to reflect spatial differences of vegetation in the study area. By analyzing the correlation between environmental monitoring different ecological zoning EVI Change and meteorological factors, vegetation management and decision-making control data reference and theoretical basis.
Based on crop blooming time, the early and late infer the growth of crops grown:
By-pixel basis for growth in each crop growth cycle peak time, infer sowing and harvest crops, and to determine the growth of crops early and late growth. When crop planting, weeding tillage usually requires, at this time little vegetation cover, EVI index values usually reaches a very low value. With the emergence of the crop and gradually grow, EVI timing curve will rise rapidly until the peak to peak growth. After the peak crop growth, into the harvest period, usually after crop harvest surface reproduce a bare state, EVI timing curve decreases rapidly. Methods determine the seeding and harvest crops there are many, as can be inferred sowing and harvesting of crops based on time-series graph valley EVI appear before and after the peak growth of crops. However, due to the interference of remote sensing image noise, valley EVI timing curve appears there is a lot of uncertainty, the direct impact on the accuracy of crop sowing and harvest acquired. For simplicity, in this patent, the time of occurrence of crops growing vigorously pushed forward 70 days, as determined sowing crops and crop sowing blooming period of time to crop growth stage is determined. At the same time the emergence of the crop blooming time pushed back 50 days, determined to harvest crops, and thus the growth of crops to peak harvest period of time determined as the late crop growth.
Prior knowledge, high spatial resolution remote sensing in the fusion of the image, for four categories of ground targets (water bodies, buildings and roads, woodland, bare soil), select 5-10 typical area;
according to Agriculture cultivation characteristics for different crops of cultivated land (rice, wheat, vegetables, cotton and the like), the same selected region typically 5-10;
Construction spectral feature space and feature different types of spectral characteristics of different wavelength bands through the typical gradation expression region vector
Multispectral data using continuous high time resolution, the values are typically calculated EVI region nearly 10 years, 10 years to extract the original time series EVI typical area.
A one-dimensional Gaussian filter to smooth it EVI original time series feature draw different types of EVI time series curve.
Combination of different plants and crops (poplar, pine, rice, wheat, vegetables, weeds, etc.) characteristic phenological obtain time series feature EVI different types of vegetation cover.
Remote Sensing crop problem faced:
(1) identify the major crops is often necessary to use the mask data arable land, since the land use / cover timeliness and accuracy of data and other aspects, inevitably leading to errors, and thus directly affect sensing the crop area estimation accuracy;
(2) the growth of crops due to the weather during frequent cloud, a number of remote sensing period inevitably affected by clouds, resulting in exponential series graph sensing interference, multi-stage or time series of remote sensing classification methods challenges.